Multiparametric MRI radiomics predicts PTEN mutation status in gliomas with AUC 0.874

An explainable MRI radiomics model predicted PTEN mutation status in adult gliomas (AUC 0.874, 95% CI 0.808–0.936, sensitivity 90.9%, specificity 71.7%) in a retrospective study of 195 patients, but limited PPV (PR-AUC 0.211) and external validation needed.

• The radiomic score remained independently associated with PTEN positivity after adjustment for age, MRI timing, and treatment exposure (OR per SD range 1.57–1.62, p≤0.005).
• The study was limited by the very small PTEN-positive cohort (n=11 of 195), severe class imbalance, and absence of external validation, making the findings exploratory.
• Most influential features were wavelet- and LoG-filtered GLCM and NGTDM texture descriptors from T1CE, T2, and FLAIR, suggesting a heterogeneous enhancement architecture phenotype.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.